无训练多信号融合，提升湍流下动态目标分割精度
Turbulence-Robust Dynamic Object Segmentation with Multi-Signal Priors and SAM2 Refinement
- 融合运动、语义、背景异常等先验，不需训练
- 在湍流场景中达0.425 mIoU与0.457 mDice
- 适合实时部署，适用于无标注湍流视频
本技术报告介绍了我们针对CVPR 2026 UG2+挑战赛第三赛道「湍流中的动态目标分割（DOST）」的解决方案。提出一种无需训练的多信号分割流程，结合预训练运动估计、自监督语义先验、无训练背景异常建模、手动校准的提议融合以及SAM2掩码优化。方法采用RAFT获取稠密运动响应，DINOv2生成语义目标性先验，ViBe实现无训练背景建模，预训练SAM2完成框提示掩码精修。系统全程运行于推理模式，避免端到端网络训练。该设计适用于湍流场景——大气扰动导致伪运动、模糊和目标间歇可见，单一运动线索不可靠。最终提交结果由官方排行榜评估，mIoU为0.425041，mDice为0.457206。由于未进行任务特定训练或微调，未来引入更强的时间关联学习、自适应提议选择或任务适配可能进一步提升性能。
原文摘要 · Abstract (English)
This technical report presents our solution for the CVPR 2026 UG2+ Challenge Track 3: Dynamic Object Segmentation in Turbulence (DOST). We design a training-free multi-signal segmentation pipeline that combines pretrained motion estimation, self-supervised semantic priors, background anomaly modeling, manually calibrated proposal fusion, and SAM2-based mask refinement. The method uses RAFT for dense motion responses, DINOv2 for semantic objectness priors, ViBe for training-free background modeling, and pretrained SAM2 for box-prompt mask refinement. Instead of optimizing an end-to-end segmentation network, our system operates entirely in inference mode. This design is suitable for the DOST setting, where severe atmospheric turbulence produces pseudo-motion, blur, and intermittent target visibility, making a single motion cue unreliable. The final submitted masks are evaluated by the official leaderboard, which reports 0.425041 mIoU and 0.457206 mDice. Since no task-specific model training or fine-tuning is performed, stronger learned temporal association, adaptive proposal selection, or task-specific adaptation may further improve the system.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。